Fusion Forum

The Cancer Cure Got Reassigned to the Chatbot

Every major AI lab promised the end of disease. Then Google disbanded its Nobel-winning AlphaFold team, and a …

A Government Wrote the Responsible-AI Manual. It Skips the Breach That Just Happened Twice.

A government published a responsible-AI readiness manual in October 2025. It governs a model that obeys you. Two …

A Bridge to a Bank That Already Rejected You

The reports diagnosed the AI-jobs crisis and skipped the method. Now the CEOs are narrating it, and every …

The "Contained" AI Breach Just Grew to Five Victims

Anthropic's AI Agent Security Manual Skips the Attack That Keeps Winning

Zero Trust for AI Agents gets the infrastructure right and underweights the way real agents get owned. If …

Recognizing Evidence, Not Conclusions

Singapore Wrote the Best AI Rulebook on the Planet. One Word Is Missing.

Who Evaluates the Evaluators?

Independence and competence criteria for third-party evaluation of frontier AI models.

A Framework of Trust That Holds as Systems Learn

France can approve a version. The system it receives keeps learning. The mechanism governing that change decides whether …

Your ATS is Working as Designed

If you run hiring, your ATS is quietly rejecting qualified people by design, not by accident. The data …

OpenAI's AI Escaped the Lab and Robbed a Partner. This Was Never a One-Off.

Stanford Told Congress What's Broken. Nobody Told Congress How to Check.

In 2024 Stanford HAI & Black in AI handed Congress a rigorous warning about how AI harms Black …